Investigation of vortex-induced vibrations to compare Reynolds-averaged Navier-Stokes and detached-eddy simulations

ifasd2024 Tracking Number 214

The increasing power of high-performance computing leads to a more efficient use of scale resolving CFD. The effects of these methods on the aeroelastic behaviour are not known well yet. The Aim is to analyse an aeroelastic test case for scale-resolved numerical flow simulation. Several two-dimensional Reynolds-averaged Navier-Stokes (RANS) and three-dimensional delayed-detached-eddy simulations (DDES) are performed with the NACA0021 airfoil at high angles of attack (AoA) of 60° and 70° at Mach 0.1 and a Reynolds number of 2.7×〖10〗^5 to investigate vortex-induced vibration (ViV). The focus is on the difference between RANS and scale-resolving methods and the analysis of the turbulence in the wake behind the wing. A rigid body test case is used as a basis. The flow behind NACA0021 shows a strong vortex system in the wake with a dominant frequency for this rigid body case. Fluid-Structure Interaction (FSI) and forced motion simulations are performed on this basis. A grid study for the RANS- and DDES-simulation was performed for for the rigid airfoil at 60° AoA. At 70° AoA, both simulation methods show a strong lock-in behaviour with a large influence on the turbulent kinetic energy in the resolved wake turbulence. Without lock-in, this energy is about ten times higher than with lock-in. The correlation of the x-velocity in the turbulent wake over the span is also investigated. Outside the lock-in region, there is no effect of motion on the span correlation. Inside the lock-in region, the correlation increases significantly in the spanwise direction. Compared to the RANS simulation, the frequency range of the lock-in is slightly smaller for DDES. The vortex-induced vibration's maximum amplitude is almost the same in both simulations. In the rigid case, there is a significant difference between RANS and DES in terms of integral forces. This difference decreases significantly when the profile is set in motion. Remarkedly there is no significant difference between RANS and DES in terms of the integral forces and the ViV amplitude and frequency on the oscillating airfoil under lock-in conditions.